Invasive annual grasses, such as medusahead, can reduce forage production capacity and interfere with revegetation projects in California rangelands. Because of the taxonomic similarity to other more desirable grasses, achieving selective control of invasive annual grasses can be difficult. In selectivity trials conducted in Yolo and Siskiyou counties, CA, the herbicide imazapic gave control of many nonnative annual grasses yet provided some level of selectivity to specific perennial grasses used in revegetation projects throughout the western United States. The selectivity difference between newly seeded perennial and annual grasses was greater with PRE applications than with POST treatments. Both perennial and annual grasses within the tribe Hordeae were more tolerant to imazapic than other grass species. In addition, field experiments were conducted at three sites in northern California (Yuba, Yolo, and Lassen counties) and one in southern Oregon (Lake County) to test the response of imazapic to varying management conditions. Imazapic was applied PRE in fall (and also spring in Lake County) at rates from 35 to 210 g/ha on undisturbed rangeland, in comparison with rangeland cleared of standing plant material and thatch by either tillage, mowing and raking, or burning. Imazapic generally showed enhanced weed control when applied following disturbance. Rates as low as 70 g/ha, if combined with thatch removal, provided significant suppression of medusahead. In addition, disturbance alone generally reduced medusahead cover in the following year. Although imazapic showed potential for control of medusahead and other annual grasses, its selectivity window was relatively narrow.